Research Article: Quality Improvement Initiatives Need Rigorous Evaluation: The Case of Pressure Ulcers

Date Published: August 30, 2016

Publisher: SAGE Publications

Author(s): Richard F. Averill, John S. Hughes, Richard L. Fuller, Norbert I. Goldfield.


The Partnership for Patients (PfP) and the Agency for Healthcare Research and Quality (AHRQ) have reported a 23.5% decline in hospital-acquired pressure ulcers (HAPU) over 4 years resulting in a cumulative cost savings of more than $10 billion and 49 000 averted deaths, claiming that this significant decline may have been spurred in part by Medicare payment incentives associated with severe (stage 3 or 4) HAPUs. Hospitals with a high rate of severe HAPUs have a payment penalty imposed, creating a financial disincentive to report severe HAPUs, possibly contributing to the magnitude of the reported decline. Despite the financial disincentive to report, the number of severe HAPUs found in claims data over the corresponding 4-year period did not decline but instead remained unchanged. The results from claims data, combined with some flaws in estimating HAPUs, call into question the validity of the decline in HAPUs reported by PfP and AHRQ.

Partial Text

Pressure ulcers are a localized injury to the skin and/or underlying tissue caused by pressure on the skin (sometimes referred to as bed sores). In administrative claims data, pressure ulcers are reported as stage 1 through 4 (International Classification of Disease, Ninth Revision, Clinical Modification [ICD-9-CM] codes 70721-70724, respectively) or, more often, without a specified stage (ICD-9-CM codes 70700-70709, 70720, 70725). Stage 3 (full thickness tissue loss) and stage 4 (full thickness tissue loss with exposed bone, tendons, or muscle) are the most severe pressure ulcers. With the implementation of Medicare Severity–Diagnosis Related Groups (MS-DRGs) in 2008, stage 3 and stage 4 pressure ulcers are considered to be a major complication or comorbidity. The presence of a secondary diagnosis that is classified as a major complication or comorbidity will often significantly increase payment under MS-DRGs. Beginning in January of 2008, hospitals were required to report in their administrative claims data whether or not each diagnosis was present on admission (POA). The POA designation allowed for the identification of hospital-acquired complications, thereby enabling the implementation of 2 policies affecting the payment of HAPUs:

The Medicare Provider Analysis and Review inpatient hospital data for FYs 2010 to 2014 (October 1, 2009, through September 30, 2014), based on computerized hospital claims data, was used to calculate the observed rates of stage 3 and stage 4 HAPUs. As shown in Table 1, the observed rate of stage 3 and stage 4 HAPUs did not decline as expected but was essentially flat from 2010 through 2014. If there had been a 23.5% decline in stage 3 and stage 4 HAPUs, the 21 993 stage 3 and stage 4 HAPUs in 2010 would have decreased by 5168 to 16 825 in 2014. In light of the financial disincentive to report stage 3 and stage 4 HAPUs, it is difficult to reconcile a lack of any actual decline in the stage 3 and stage 4 HAPUs during the 4-year period in which AHRQ and PfP reported a dramatic decline in HAPUs across all stages. In contrast, the number of stage 1, stage 2, and unstaged HAPUs increased from 83 545 in 2010 to 132 062 (a rate increase from 0.57% to 0.88%).

The estimated 23.5% decline in HAPUs and the associated cost savings of more than $10 billion between 2010 and 2014 were derived from the Medicare Patient Safety Monitoring System (MPSMS) chart reviews of a limited subset of Medicare patients varying between 18 000 and 33 000 patients in a given year.6 It could be argued that chart reviews produce more accurate data on HAPUs than is reported in claims data. Although there is some merit to that argument for HAPUs that do not impact payment (stage 1, stage 2, and stage unspecified HAPUs), it is unlikely to be true for stage 3 and stage 4 HAPUs that directly impact payment. If the decline in stage 3 and stage 4 HAPUs derived from the MPSMS chart reviews was accurate, the lack of an actual decline in the claims data would imply that hospitals are systematically overreporting stage 3 and stage 4 HAPUs. Overreporting of stage 3 and stage 4 HAPUs by hospitals seems very unlikely in light of the financial disincentive to report them. As a result, the decline in HAPU rates reported by AHRQ and PfP are at least counterintuitive and at best questionable.

The 23.5% decrease in HAPUs reported by AHRQ and PfP cannot be verified using the Medicare claims database. The estimated 23.5% decline in HAPUs and the associated cost savings of more than $10 billion are based on questionable assumptions used for extrapolating Medicare results to the general population. Indeed, based on claims data, the estimated cost savings are virtually zero. A thorough and valid evaluation of the AHRQ and PfP reported results is needed before using the results as the basis for future policy decisions.




Leave a Reply

Your email address will not be published.